#!/usr/bin/env python
# Created by "Thieu" at 11:16, 18/03/2020 ----------%
# Email: nguyenthieu2102@gmail.com %
# Github: https://github.com/thieu1995 %
# --------------------------------------------------%
import numpy as np
from mealpy.optimizer import Optimizer
[docs]class DevSARO(Optimizer):
"""
The developed version: Search And Rescue Optimization (SARO)
Hyper-parameters should fine-tune in approximate range to get faster convergence toward the global optimum:
+ se (float): [0.3, 0.8], social effect, default = 0.5
+ mu (int): maximum unsuccessful search number, belongs to range: [2, 2+int(self.pop_size/2)], default = 15
Examples
~~~~~~~~
>>> import numpy as np
>>> from mealpy import FloatVar, SARO
>>>
>>> def objective_function(solution):
>>> return np.sum(solution**2)
>>>
>>> problem_dict = {
>>> "bounds": FloatVar(n_vars=30, lb=(-10.,) * 30, ub=(10.,) * 30, name="delta"),
>>> "minmax": "min",
>>> "obj_func": objective_function
>>> }
>>>
>>> model = SARO.DevSARO(epoch=1000, pop_size=50, se = 0.5, mu = 50)
>>> g_best = model.solve(problem_dict)
>>> print(f"Solution: {g_best.solution}, Fitness: {g_best.target.fitness}")
>>> print(f"Solution: {model.g_best.solution}, Fitness: {model.g_best.target.fitness}")
"""
def __init__(self, epoch: int = 10000, pop_size: int = 100, se: float = 0.5, mu: int = 15, **kwargs: object) -> None:
"""
Args:
epoch (int): maximum number of iterations, default = 10000
pop_size (int): number of population size, default = 100
se (float): social effect, default = 0.5
mu (int): maximum unsuccessful search number, default = 15
"""
super().__init__(**kwargs)
self.epoch = self.validator.check_int("epoch", epoch, [1, 100000])
self.pop_size = self.validator.check_int("pop_size", pop_size, [5, 10000])
self.se = self.validator.check_float("se", se, (0, 1.0))
self.mu = self.validator.check_int("mu", mu, [2, 2+int(self.pop_size/2)])
self.set_parameters(["epoch", "pop_size", "se", "mu"])
self.sort_flag = True
[docs] def initialize_variables(self):
self.dyn_USN = np.zeros(self.pop_size)
[docs] def initialization(self):
if self.pop is None:
self.pop = self.generate_population(2 * self.pop_size)
else:
self.pop = self.pop + self.generate_population(self.pop_size)
[docs] def amend_solution(self, solution: np.ndarray) -> np.ndarray:
condition = np.logical_and(self.problem.lb <= solution, solution <= self.problem.ub)
rand_pos = self.generator.uniform(self.problem.lb, self.problem.ub)
return np.where(condition, solution, rand_pos)
[docs] def evolve(self, epoch):
"""
The main operations (equations) of algorithm. Inherit from Optimizer class
Args:
epoch (int): The current iteration
"""
pop_x = [agent.copy() for agent in self.pop[:self.pop_size]]
pop_m = [agent.copy() for agent in self.pop[self.pop_size:]]
pop_new = []
for idx in range(self.pop_size):
## Social Phase
k = self.generator.choice(list(set(range(0, 2 * self.pop_size)) - {idx}))
sd = pop_x[idx].solution - self.pop[k].solution
#### Remove third loop here, also using random flight back when out of bound
pos_new_1 = self.pop[k].solution + self.generator.uniform() * sd
pos_new_2 = pop_x[idx].solution + self.generator.uniform() * sd
condition = np.logical_and(self.generator.uniform(0, 1, self.problem.n_dims) < self.se, self.pop[k].target.fitness < pop_x[idx].target.fitness)
pos_new = np.where(condition, pos_new_1, pos_new_2)
pos_new = self.correct_solution(pos_new)
agent = self.generate_empty_agent(pos_new)
pop_new.append(agent)
if self.mode not in self.AVAILABLE_MODES:
pop_new[-1].target = self.get_target(pos_new)
pop_new = self.update_target_for_population(pop_new)
for idx in range(self.pop_size):
if self.compare_target(pop_new[idx].target, pop_x[idx].target, self.problem.minmax):
pop_m[self.generator.integers(0, self.pop_size)] = pop_x[idx].copy()
pop_x[idx] = pop_new[idx].copy()
self.dyn_USN[idx] = 0
else:
self.dyn_USN[idx] += 1
pop = pop_x.copy() + pop_m.copy()
pop_new = []
for idx in range(self.pop_size):
## Individual phase
k1, k2 = self.generator.choice(list(set(range(0, 2 * self.pop_size)) - {idx}), 2, replace=False)
#### Remove third loop here, and flight back strategy now be a random
pos_new = self.g_best.solution + self.generator.uniform() * (pop[k1].solution - pop[k2].solution)
pos_new = self.correct_solution(pos_new)
agent = self.generate_empty_agent(pos_new)
pop_new.append(agent)
if self.mode not in self.AVAILABLE_MODES:
pop_new[-1].target = self.get_target(pos_new)
pop_new = self.update_target_for_population(pop_new)
for idx in range(0, self.pop_size):
if self.compare_target(pop_new[idx].target, pop_x[idx].target, self.problem.minmax):
pop_m[self.generator.integers(0, self.pop_size)] = pop_x[idx].copy()
pop_x[idx] = pop_new[idx].copy()
self.dyn_USN[idx] = 0
else:
self.dyn_USN[idx] += 1
if self.dyn_USN[idx] > self.mu:
pop_x[idx] = self.generate_agent()
self.dyn_USN[idx] = 0
self.pop = pop_x + pop_m
[docs]class OriginalSARO(DevSARO):
"""
The original version of: Search And Rescue Optimization (SARO)
Links:
1. https://doi.org/10.1155/2019/2482543
Hyper-parameters should fine-tune in approximate range to get faster convergence toward the global optimum:
+ se (float): [0.3, 0.8], social effect, default = 0.5
+ mu (int): [10, 20], maximum unsuccessful search number, default = 15
Examples
~~~~~~~~
>>> import numpy as np
>>> from mealpy import FloatVar, SARO
>>>
>>> def objective_function(solution):
>>> return np.sum(solution**2)
>>>
>>> problem_dict = {
>>> "bounds": FloatVar(n_vars=30, lb=(-10.,) * 30, ub=(10.,) * 30, name="delta"),
>>> "minmax": "min",
>>> "obj_func": objective_function
>>> }
>>>
>>> model = SARO.OriginalSARO(epoch=1000, pop_size=50, se = 0.5, mu = 50)
>>> g_best = model.solve(problem_dict)
>>> print(f"Solution: {g_best.solution}, Fitness: {g_best.target.fitness}")
>>> print(f"Solution: {model.g_best.solution}, Fitness: {model.g_best.target.fitness}")
References
~~~~~~~~~~
[1] Shabani, A., Asgarian, B., Gharebaghi, S.A., Salido, M.A. and Giret, A., 2019. A new optimization
algorithm based on search and rescue operations. Mathematical Problems in Engineering, 2019.
"""
def __init__(self, epoch: int = 10000, pop_size: int = 100, se: float = 0.5, mu: int = 15, **kwargs: object) -> None:
"""
Args:
epoch (int): maximum number of iterations, default = 10000
pop_size (int): number of population size, default = 100
se (float): social effect, default = 0.5
mu (int): maximum unsuccessful search number, default = 15
"""
super().__init__(epoch, pop_size, se, mu, **kwargs)
[docs] def evolve(self, epoch):
"""
The main operations (equations) of algorithm. Inherit from Optimizer class
Args:
epoch (int): The current iteration
"""
pop_x = [agent.copy() for agent in self.pop[:self.pop_size]]
pop_m = [agent.copy() for agent in self.pop[self.pop_size:]]
pop_new = []
for idx in range(self.pop_size):
## Social Phase
k = self.generator.choice(list(set(range(0, 2 * self.pop_size)) - {idx}))
sd = pop_x[idx].solution - self.pop[k].solution
j_rand = self.generator.integers(0, self.problem.n_dims)
r1 = self.generator.uniform(-1, 1)
pos_new = pop_x[idx].solution.copy()
for j in range(0, self.problem.n_dims):
if self.generator.uniform() < self.se or j == j_rand:
if self.compare_target(self.pop[k].target, pop_x[idx].target, self.problem.minmax):
pos_new[j] = self.pop[k].solution[j] + r1 * sd[j]
else:
pos_new[j] = pop_x[idx].solution[j] + r1 * sd[j]
if pos_new[j] < self.problem.lb[j]:
pos_new[j] = (pop_x[idx].solution[j] + self.problem.lb[j]) / 2
if pos_new[j] > self.problem.ub[j]:
pos_new[j] = (pop_x[idx].solution[j] + self.problem.ub[j]) / 2
pos_new = self.correct_solution(pos_new)
agent = self.generate_empty_agent(pos_new)
pop_new.append(agent)
if self.mode not in self.AVAILABLE_MODES:
pop_new[-1].target = self.get_target(pos_new)
pop_new = self.update_target_for_population(pop_new)
for idx in range(0, self.pop_size):
if self.compare_target(pop_new[idx].target, pop_x[idx].target, self.problem.minmax):
pop_m[self.generator.integers(0, self.pop_size)] = pop_x[idx].copy()
pop_x[idx] = pop_new[idx].copy()
self.dyn_USN[idx] = 0
else:
self.dyn_USN[idx] += 1
## Individual phase
pop = pop_x.copy() + pop_m.copy()
pop_new = []
for idx in range(0, self.pop_size):
k, m = self.generator.choice(list(set(range(0, 2 * self.pop_size)) - {idx}), 2, replace=False)
pos_new = pop_x[idx].solution + self.generator.uniform() * (pop[k].solution - pop[m].solution)
for j in range(0, self.problem.n_dims):
if pos_new[j] < self.problem.lb[j]:
pos_new[j] = (pop_x[idx].solution[j] + self.problem.lb[j]) / 2
if pos_new[j] > self.problem.ub[j]:
pos_new[j] = (pop_x[idx].solution[j] + self.problem.ub[j]) / 2
pos_new = self.correct_solution(pos_new)
agent = self.generate_empty_agent(pos_new)
pop_new.append(agent)
if self.mode not in self.AVAILABLE_MODES:
pop_new[-1].target = self.get_target(pos_new)
pop_new = self.update_target_for_population(pop_new)
for idx in range(0, self.pop_size):
if self.compare_target(pop_new[idx].target, pop_x[idx].target, self.problem.minmax):
pop_m[self.generator.integers(0, self.pop_size)] = pop_x[idx]
pop_x[idx] = pop_new[idx].copy()
self.dyn_USN[idx] = 0
else:
self.dyn_USN[idx] += 1
if self.dyn_USN[idx] > self.mu:
pop_x[idx] = self.generate_agent()
self.dyn_USN[idx] = 0
self.pop = pop_x + pop_m